[BioPython] Lowess function for nonparametric regression

[Michiel Jan Laurens de Hoon]

Brad Chapman wrote:
> The Lowess.py under Bio makes the most sense under the current
> scheme, but I guess I'm not sure if that is the overall best thing.
> Do you think it would be more useful to have a "Stats" module or
> something which would collect some of these various useful
> statistics bits into an easier to find place then scattered
> throughout Bio? I don't really think we need a deep directory
> structure, but maybe something analogous to Bio/SeqUtils. 
> 
> On the other hand, I guess I'm not sure if this would lead to moving
> a lot of modules under Bio to there, which might be too much of a
> mess.

I had a look at the other modules under Bio. There are a couple of supervised 
learning routines there, but they wouldn't fit well as part of a Bio.Statistics 
module. However, there are a couple of other statistics-related routines that 
would be nice to have in Biopython (e.g. routines to estimate probability 
density functions or alternative non-parametric regression models). So I would 
argue for a Bio.Statistics module containing the Lowess function for now and 
hopefully other routines later, but keep the existing modules under Bio where 
they are.

--Michiel.

-- 
Michiel de Hoon, Assistant Professor
University of Tokyo, Institute of Medical Science
Human Genome Center
4-6-1 Shirokane-dai, Minato-ku
Tokyo 108-8639
Japan
http://bonsai.ims.u-tokyo.ac.jp/~mdehoon